ARIMA models of the price level: an assessment of the multilevel adaptive learning process in the USA

Article Abstract:

The three-level adaptive expectations model of Jacobs and Jones is shown to be an ARIMA process for the observed and expected price level, thus allowing a clearer estimate of the model's parameters and validity. This is achieved through a comparison with the best-fitting ARIMA model, which is proven to be a restricted version of a two-level Jacobs and Jones model. This suggests a simple adaptive rule in the inflation rate, while the accuracy and efficiency are also demonstrated to be adequate. However, the best-fit ARIMA model does not dovetail with any version of the Jacobs and Jones model.

Economic forecasting, Maximum likelihood estimates (Statistics), Maximum likelihood (Statistics)

---

Forecasting and analyzing economic activity with coincident and leading indexes: the case of Connecticut

Article Abstract:

Coincident and leading employment indexes were developed for the state of Connecticut as a policy decision-making tool for both businessmen and state officials. The indexes were developed by using data from 1969 to 1991 which was divided into component series and subjected to symmetrical percentage changes computations. These indexes were then subjected to a univariate and three multivariate time series forecast tests. The deficiencies of the indexes include the over-prediction of the coincident index which may be due to the accuracy of measures of employment.

Administration of General Economic Programs, Economic Planning & Research, Economic conditions, Economic indicators, Employment forecasting

---

A BVAR model for the Connecticut economy

Article Abstract:

A Bayesian vector autoregressive (BVAR) characterization of the Connecticut economy was derived to facilitate the prediction of the state's unemployment rate, nonagricultural employment rate, real personal income and authorized housing permits. The model, which encapsulates both state and national variables, was also evaluated against the unrestricted VAR model and the univariate ARIMA model. The results indicated the higher accuracy of the BVAR model with respect to short- and long-term out-of-sample forecasts.

Economic aspects, Bayesian statistical decision theory, Bayesian analysis, Autoregression (Statistics), Prediction theory

Similar abstracts:

• Abstracts: A comparison of approximate Bayesian forecasting methods for non-Gaussian time series. ARMA models and the Box-Jenkins methodology
• Abstracts: Pathological Outcomes of Observational Learning. Mechanism Design with Collusion and Correlation
• Abstracts: A note on Abreu-Matsushima mechanisms. Virtual implementation in iteratively undominated strategies: complete information

This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.